{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "1d42f85e-a931-4942-aa6e-b833291f102a",
   "metadata": {},
   "source": [
    "<a href=\"https://colab.research.google.com/github/meta-llama/llama-recipes/blob/main/recipes/use_cases/agents/langchain/langgraph-agent.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "debae1a8-61e8-448c-a2f5-550ca9908366",
   "metadata": {},
   "outputs": [],
   "source": [
    "! pip install -U langchain_groq langchain langchain_community sentence_transformers tavily-python tiktoken langchainhub chromadb langgraph"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ccef71bf-6f5d-4cce-80a2-1bb848638832",
   "metadata": {},
   "source": [
    "# LangGraph agent with Llama 3\n",
    "\n",
    "Previously, we showed how to build simple agents with the LangChain [tool calling agent](https://python.langchain.com/docs/modules/agents/agent_types/tool_calling/).\n",
    "\n",
    "However, we see that this can fail in some cases - e.g., if the LLM is not trained or prompted to use tools reliabily.\n",
    "\n",
    "LangGraph is a framework from the LangChain team that can be used to implement core [agent](https://www.deeplearning.ai/the-batch/how-agents-can-improve-llm-performance/) [principles](https://lilianweng.github.io/posts/2023-06-23-agent/), including:\n",
    "\n",
    "- **Planning:** Break down task into smaller subgoals\n",
    "- **Memory:** Short-term (e.g., chat history) and / or long-term (e.g., vectorstore)\n",
    "- **Tool use:** Tools (e.g., web search)\n",
    "- **Multi-agent:** Several agents collaborating\n",
    "\n",
    "With **LangGraph**: \n",
    "- **Planning:** The user will lay out a control flow ahead of time as a graph\n",
    "- **Memory:** The graph state persists relevant information (e.g., documents, question) across the life of the agent\n",
    "- **Tool use:** Each graph node performs a specific task that modifies state, which can include tool use\n",
    "\n",
    "Let's walk through the basic ideas in the [previously shown tool-calling-agent](https://python.langchain.com/docs/modules/agents/agent_types/tool_calling/) using LangGraph."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "89b63563-4438-48d4-9d29-060089580601",
   "metadata": {},
   "source": [
    "## Search agent\n",
    "\n",
    "Let's create a simple agent that can execute web search to supplement the LLM's knowledge.\n",
    "\n",
    "Let's use [Tavily](https://tavily.com/#api) for web search."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6b7b10ae-5213-4992-b08e-051a21024392",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "\n",
    "os.environ['TAVILY_API_KEY'] = 'YOUR_TAVILY_API_KEY'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b7f92fd1-d985-42e8-9ba1-6b193d2d452b",
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain_community.tools.tavily_search import TavilySearchResults\n",
    "web_search_tool = TavilySearchResults(k=3)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "10c8631d-713a-4e6b-b932-046f3214501b",
   "metadata": {},
   "source": [
    "Let's also use RAG on search results.\n",
    "\n",
    "We'll use [Groq](https://groq.com/) to access Llama 3 - to get a free Groq API key, sign in with your github or gmail account [here](https://console.groq.com/)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6b4f3fd9-54a3-4c4b-aece-f12d2c6cce64",
   "metadata": {},
   "outputs": [],
   "source": [
    "os.environ['GROQ_API_KEY'] = 'YOUR_GROQ_API_KEY'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "588315b6-8f31-4f6b-891f-7083b3caaf70",
   "metadata": {},
   "outputs": [],
   "source": [
    "### Generate\n",
    "\n",
    "from langchain import hub\n",
    "from langchain_groq import ChatGroq\n",
    "from langchain_core.output_parsers import StrOutputParser\n",
    "\n",
    "# Prompt\n",
    "prompt = hub.pull(\"rlm/rag-prompt\")\n",
    "prompt.pretty_print()\n",
    "\n",
    "# LLM\n",
    "llm = ChatGroq(temperature=0, model=\"llama3-70b-8192\")\n",
    "\n",
    "# Post-processing\n",
    "def format_docs(docs):\n",
    "    return \"\\n\\n\".join(doc.page_content for doc in docs)\n",
    "\n",
    "# Chain\n",
    "rag_chain = prompt | llm | StrOutputParser()"
   ]
  },
  {
   "attachments": {
    "de16ec74-7f40-4c92-983e-367786cde64d.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "id": "386780a6-7b43-4c3e-9857-97f7d3921448",
   "metadata": {},
   "source": [
    "Here is the simple graph:\n",
    "\n",
    "![Screenshot 2024-05-03 at 11.53.07 AM.png](attachment:de16ec74-7f40-4c92-983e-367786cde64d.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ec515683-bcfa-4173-b2c1-a70532b15f34",
   "metadata": {},
   "source": [
    "The graph state is our short-term memory.\n",
    "\n",
    "We'll use it to hold any information that we want our agent to use."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "858092fd-09d9-43aa-a62d-90b57d8ab518",
   "metadata": {},
   "outputs": [],
   "source": [
    "from typing_extensions import TypedDict\n",
    "from typing import List\n",
    "\n",
    "class GraphState(TypedDict):\n",
    "    \"\"\"\n",
    "    Represents the state of our graph.\n",
    "\n",
    "    Attributes:\n",
    "        question: question\n",
    "        answer: answer\n",
    "        web_search: Tavily web search results\n",
    "    \"\"\"\n",
    "    question : str\n",
    "    web_search : str\n",
    "    answer : str"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6936b541-bd70-4e4d-b46f-7a90b9797d66",
   "metadata": {},
   "source": [
    "We define a graph nodes that performs web search and RAG."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "998bb0ef-68b5-4f13-8fb4-ac52da2d70ee",
   "metadata": {},
   "outputs": [],
   "source": [
    "def web_search(state):\n",
    "    \"\"\"\n",
    "    Web search based based on the question\n",
    "\n",
    "    Args:\n",
    "        state (dict): The current graph state\n",
    "\n",
    "    Returns:\n",
    "        state (dict): Appended web results to web_search\n",
    "    \"\"\"\n",
    "\n",
    "    print(\"---WEB SEARCH---\")\n",
    "    question = state[\"question\"]\n",
    "    docs = web_search_tool.invoke({\"query\": question})\n",
    "    web_search = \"\\n\".join([d[\"content\"] for d in docs])\n",
    "    return {\"web_search\": web_search, \"question\": question}\n",
    "\n",
    "def generate(state):\n",
    "    \"\"\"\n",
    "    Generate answer using RAG on web search\n",
    "\n",
    "    Args:\n",
    "        state (dict): The current graph state\n",
    "\n",
    "    Returns:\n",
    "        state (dict): New key added to state, answer, that contains LLM generation\n",
    "    \"\"\"\n",
    "    print(\"---GENERATE---\")\n",
    "    question = state[\"question\"]\n",
    "    web_search = state[\"web_search\"]\n",
    "    \n",
    "    # RAG generation\n",
    "    generation = rag_chain.invoke({\"context\": web_search, \"question\": question})\n",
    "    return {\"web_search\": web_search, \"question\": question, \"answer\": generation}"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0862ee2e-3ed9-4c54-a075-180b74a553af",
   "metadata": {},
   "source": [
    "We define a graph."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "91e6bb3d-ccec-41be-b864-97594a3a8f22",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Register the nodes\n",
    "from langgraph.graph import END, StateGraph\n",
    "workflow = StateGraph(GraphState)\n",
    "workflow.add_node(\"websearch\", web_search) # web search\n",
    "workflow.add_node(\"generate\", generate) # rag"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4a8c251d-5850-4f09-a0d7-2b7c06b1847d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Build graph\n",
    "workflow.set_entry_point(\"websearch\")\n",
    "workflow.add_edge(\"websearch\", \"generate\")\n",
    "workflow.add_edge(\"generate\", END)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ac3b61a0-c666-4f6c-9037-d7bc9446d88d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Compile and run\n",
    "app = workflow.compile()\n",
    "app.invoke({\"question\":\"what is llama3?\"})"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2ac2b01d-4f23-4caa-ac8d-de5edfcbd034",
   "metadata": {},
   "source": [
    "Let's review\n",
    "\n",
    "- **Planning:** We laid out the control flow ahead of time with search and RAG\n",
    "- **Memory:** Use graph state to persist relevant information (e.g., search results, question)\n",
    "- **Tool use:** We defined a node that invoked the search tool\n",
    "\n",
    "Here is the trace: \n",
    "\n",
    "https://smith.langchain.com/public/1ec212d9-2979-4e6a-af41-c0d7cf39f18b/r"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "678fb518-a68e-4b00-b61d-406b3ee15f77",
   "metadata": {},
   "source": [
    "## Search and Retrieval Agent\n",
    "\n",
    "We can give our LangGraph agent multiple tools.\n",
    "\n",
    "For example, let's assume we want to perform retrieval and web search.\n",
    "\n",
    "We'll use retrieval if the question relates to LangSmith."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5f9264f3-39a9-4c42-b738-a0c56897b576",
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain_community.document_loaders import WebBaseLoader\n",
    "from langchain_community.vectorstores import Chroma\n",
    "from langchain_community.embeddings import HuggingFaceEmbeddings\n",
    "from langchain_text_splitters import RecursiveCharacterTextSplitter\n",
    "\n",
    "# Retriever\n",
    "loader = WebBaseLoader(\"https://docs.smith.langchain.com/overview\")\n",
    "docs = loader.load()\n",
    "documents = RecursiveCharacterTextSplitter(\n",
    "    chunk_size=1000, chunk_overlap=200\n",
    ").split_documents(docs)\n",
    "vector = Chroma.from_documents(documents, HuggingFaceEmbeddings())\n",
    "retriever = vector.as_retriever()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9bdfe2ae-d4ed-4fee-a7be-a03c3a5ab198",
   "metadata": {},
   "source": [
    "We'll need a way to route questions between vectorstore and web search."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b6e4f57f-3953-4117-aa1d-9e08184466cc",
   "metadata": {},
   "outputs": [],
   "source": [
    "### Router\n",
    "\n",
    "from typing import Literal\n",
    "\n",
    "from langchain_core.prompts import ChatPromptTemplate\n",
    "from langchain_core.pydantic_v1 import BaseModel, Field\n",
    "from langchain_groq import ChatGroq\n",
    "\n",
    "# Data model\n",
    "class RouteQuery(BaseModel):\n",
    "    \"\"\"Route a user query to the most relevant datasource.\"\"\"\n",
    "\n",
    "    datasource: Literal[\"vectorstore\", \"web_search\"] = Field(\n",
    "        ...,\n",
    "        description=\"Given a user question choose to route it to web search or a vectorstore.\",\n",
    "    )\n",
    "\n",
    "# LLM with function call \n",
    "llm = ChatGroq(temperature=0, model=\"llama3-70b-8192\")\n",
    "structured_llm_router = llm.with_structured_output(RouteQuery)\n",
    "\n",
    "# Prompt \n",
    "system = \"\"\"You are an expert at routing a user question to a vectorstore or web search.\n",
    "The vectorstore contains information related to LangSmith. Use the vectorstore for questions \n",
    "that mention LangSmith. Otherwise, use web-search.\"\"\"\n",
    "route_prompt = ChatPromptTemplate.from_messages(\n",
    "    [\n",
    "        (\"system\", system),\n",
    "        (\"human\", \"{question}\"),\n",
    "    ]\n",
    ")\n",
    "\n",
    "question_router = route_prompt | structured_llm_router\n",
    "print(question_router.invoke({\"question\": \"Who will the Bears draft first in the NFL draft?\"}))\n",
    "print(question_router.invoke({\"question\": \"What is LangSmith?\"}))"
   ]
  },
  {
   "attachments": {
    "9818fb0f-9cbc-4260-a4ed-46634c8a1dc5.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "id": "37b652da-5fad-4c71-a20c-170323ff9e1a",
   "metadata": {},
   "source": [
    "Here is our graph flow.\n",
    "\n",
    "![Screenshot 2024-05-03 at 12.27.51 PM.png](attachment:9818fb0f-9cbc-4260-a4ed-46634c8a1dc5.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "49f74b0c-1253-4451-9e5c-f653396f6cb5",
   "metadata": {},
   "source": [
    "Set state, similar to above."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "847cfc69-9ce9-4531-94ce-fb3c6a07df21",
   "metadata": {},
   "outputs": [],
   "source": [
    "from typing_extensions import TypedDict\n",
    "from typing import List\n",
    "\n",
    "class GraphState(TypedDict):\n",
    "    \"\"\"\n",
    "    Represents the state of our graph.\n",
    "\n",
    "    Attributes:\n",
    "        question: question\n",
    "        answer: answer\n",
    "        context: Docs from web search or vectorstore\n",
    "    \"\"\"\n",
    "    question : str\n",
    "    context : str\n",
    "    answer : str"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "23c35026-d00b-4770-9d91-56e55fac7c03",
   "metadata": {},
   "source": [
    "Set nodes, similar to above."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "51ec8381-00d7-4311-b857-cf304e88f296",
   "metadata": {},
   "outputs": [],
   "source": [
    "def retrieve(state):\n",
    "    \"\"\"\n",
    "    Web search based based on the question\n",
    "\n",
    "    Args:\n",
    "        state (dict): The current graph state\n",
    "\n",
    "    Returns:\n",
    "        state (dict): Appended web results to web_search\n",
    "    \"\"\"\n",
    "\n",
    "    print(\"---RETRIEVE---\")\n",
    "    question = state[\"question\"]\n",
    "    docs = retriever.invoke(question)\n",
    "    docs_str = \"\\n\".join([d.page_content for d in docs])\n",
    "    return {\"context\": docs_str, \"question\": question}\n",
    "\n",
    "def web_search(state):\n",
    "    \"\"\"\n",
    "    Web search based based on the question\n",
    "\n",
    "    Args:\n",
    "        state (dict): The current graph state\n",
    "\n",
    "    Returns:\n",
    "        state (dict): Appended web results to web_search\n",
    "    \"\"\"\n",
    "\n",
    "    print(\"---WEB SEARCH---\")\n",
    "    question = state[\"question\"]\n",
    "    docs = web_search_tool.invoke({\"query\": question})\n",
    "    web_search = \"\\n\".join([d[\"content\"] for d in docs])\n",
    "    return {\"context\": web_search, \"question\": question}\n",
    "\n",
    "def generate(state):\n",
    "    \"\"\"\n",
    "    Generate answer using RAG on web search\n",
    "\n",
    "    Args:\n",
    "        state (dict): The current graph state\n",
    "\n",
    "    Returns:\n",
    "        state (dict): New key added to state, answer, that contains LLM generation\n",
    "    \"\"\"\n",
    "    print(\"---GENERATE---\")\n",
    "    question = state[\"question\"]\n",
    "    context = state[\"context\"]\n",
    "    \n",
    "    # RAG generation\n",
    "    generation = rag_chain.invoke({\"context\": context, \"question\": question})\n",
    "    return {\"context\": context, \"question\": question, \"answer\": generation}"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d915c3f8-1e57-4db0-a1df-824e9c3a4853",
   "metadata": {},
   "source": [
    "Now, we have one conditional edge. \n",
    "\n",
    "The output of this conditerional edge determine the next node to visit."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7e028ee1-4613-4502-8a7d-7eae306790e2",
   "metadata": {},
   "outputs": [],
   "source": [
    "### Conditional edge\n",
    "\n",
    "def route_question(state):\n",
    "    \"\"\"\n",
    "    Route question to web search or vectorstore.\n",
    "\n",
    "    Args:\n",
    "        state (dict): The current graph state\n",
    "\n",
    "    Returns:\n",
    "        str: Next node to call\n",
    "    \"\"\"\n",
    "\n",
    "    print(\"---ROUTE QUESTION---\")\n",
    "    question = state[\"question\"]\n",
    "    source = question_router.invoke({\"question\": question})  \n",
    "    if source.datasource == 'web_search':\n",
    "        print(\"---ROUTE QUESTION TO WEB SEARCH---\")\n",
    "        return \"websearch\"\n",
    "    elif source.datasource == 'vectorstore':\n",
    "        print(\"---ROUTE QUESTION TO RAG---\")\n",
    "        return \"vectorstore\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4ed0b5b0-b2d3-462c-9654-a6b6b595c2e6",
   "metadata": {},
   "source": [
    "Add nodes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ac854850-9d57-4bf0-8982-8c3eea2e15d5",
   "metadata": {},
   "outputs": [],
   "source": [
    "from langgraph.graph import END, StateGraph\n",
    "workflow = StateGraph(GraphState)\n",
    "\n",
    "# Define the nodes\n",
    "workflow.add_node(\"websearch\", web_search) # web search\n",
    "workflow.add_node(\"retrieve\", retrieve) # retrieve\n",
    "workflow.add_node(\"generate\", generate) # generatae"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7edafcc2-c815-4633-bfdc-8e0f0d140399",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Build graph\n",
    "workflow.set_conditional_entry_point(\n",
    "    route_question,\n",
    "    {   # Decision from edge -> Next node to visit\n",
    "        \"websearch\": \"websearch\",\n",
    "        \"vectorstore\": \"retrieve\",\n",
    "    },\n",
    ")\n",
    "workflow.add_edge(\"websearch\", \"generate\")\n",
    "workflow.add_edge(\"retrieve\", \"generate\")\n",
    "workflow.add_edge(\"generate\", END)\n",
    "\n",
    "# Compile\n",
    "app = workflow.compile()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fb19e045-8c4d-4d28-9f5f-7d3bc6392405",
   "metadata": {},
   "outputs": [],
   "source": [
    "app.invoke({\"question\":\"Who did the Bears draft first in the NFL draft?\"})"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bb57a811-8468-4ef2-b494-04c60cd18a56",
   "metadata": {},
   "source": [
    "Trace: \n",
    "\n",
    "https://smith.langchain.com/public/16b2ea06-5d2a-451e-b77d-7c5c0baefa04/r"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "23ea00d5-17f9-429c-9eb0-e8b3c76b68bc",
   "metadata": {},
   "outputs": [],
   "source": [
    "app.invoke({\"question\":\"how can langsmith help with testing?\"})"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9cba076a-cb39-44a4-bc81-8f0cf5fbbf9c",
   "metadata": {},
   "source": [
    "Trace: \n",
    "\n",
    "https://smith.langchain.com/public/54bc2dcf-aab7-49f5-b144-11d87f7ad799/r"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "db1c295c-d10d-4d28-8c43-776c85fadbaa",
   "metadata": {},
   "source": [
    "### Trace-offs vs Agent Executor\n",
    "\n",
    "There are trade-offs between LangGraph and Agent Executor for implementing agents.\n",
    "\n",
    "#### LangGraph\n",
    "\n",
    "`Pros -`\n",
    "\n",
    "It is highly reliable.\n",
    "\n",
    "And it easy to audit.\n",
    "\n",
    "`Cons -` \n",
    "\n",
    "LangGraph requires more code the lay out the same functionality as shown in the first Agent Executor notebook. \n",
    "\n",
    "LangGraph is less flexible because the control flow is set by the developer. It can only follow the plan set by the developer.\n",
    "\n",
    "#### Agent Executor\n",
    "\n",
    "`Pros -`\n",
    "\n",
    "Agent Executor is more flexible; the LLM can choose any tool at each step in the agent's reasoning.\n",
    "\n",
    "It also requires fewer lines to code to implement the same functionality relative to LangGraph.\n",
    "\n",
    "`Cons -`\n",
    "\n",
    "But it needs high quality tool-calling, as the LLM is responisibile for making decisions about which tool to use at each step.\n",
    "\n",
    "Also, it is harder to audit. \n",
    "\n",
    "For example, [here](https://smith.langchain.com/public/92c8157d-da30-475b-bfb7-076fd1be4377/r) is the trace of the same workflow as we did above using Agent Executor.\n",
    "\n",
    "We can see that the agent correctly decides to query the vectorstore for a question about LangSmith. And it appears to get stuck in a loop of calling various tools before crashing.\n",
    "\n",
    "Of course, this is using a non-fine-tuned (only prompting) version of Llama 3 for tool-use. But it illustates the reliability challenges with using Agent Executor."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "15960e72-5499-4b4c-b4ca-075fe3f9baca",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}